Subject: Re: Followup on Exhibit B licences
From: Rick Moen <rick@linuxmafia.com>
Date: Wed, 7 Mar 2007 01:48:53 -0800

Quoting Lawrence Rosen (lrosen@rosenlaw.com):

> An open source license cannot (should not!) contain a trademark license.

A possible compare-and-contrast item:  A gentleman named Larry Ozeran
wrote to me, this evening, to ask my view of the Medsphere Public
Licence (http://medsphere.org/license/MSPL.html) -- which task I
approached with no enthusiasm, because I expected it would be more of
the same.  It wasn't. 

Quoting Larry Ozeran (lozeran@clinicalinformatics.com):

[...]

> My question to you: is the "MSPL" license (apparently a form of
> badgeware) which they are requiring a valid open source license? Is
> there such a thing as "valid" or is the better distinction OSI
> compliant license vs. not?

[...]

I find myself pleasantly surprised and impressed by Medsphere Public
License (MSPL) Version 1.0.  Having encountered a number of other MPL
1.1 + "Exhibit B" licences, I expected the same sort of meddlesome
overreaching exemplified by the originator of such licences, SugarCRM,
Inc.  Medsphere Systems Corporation didn't do that.

MPSL's Exhibit B is a reasonable, minimal, well-worded provision
intended to ensure that the original developer (in this case,
Medsphere) continues to have publicly verifiable author credit that can
be read at runtime by all users, even on derivative works deployed
ASP-style with no distribution of source code to anyone.  The ASP
market is a challenging one for would-be publishers of open source code 
in this regard, in that -- if using a conventional copyleft or
simple-permissive open-source licence -- competitors can fully reuse
your works in their derivative work without either giving back their
changes nor leaving signs of your authorship visible to users.

Other Exhibit B sections I've seen very definitely violate OSD#10 
(the requirement for technological neutrality).  MSPL carefully avoids
that pitfall by saying the splash-screen credit must appear on the
graphical user interface screen _if any_.  

Again, in sharp contrast to other Exhibit B licences, MSPL doesn't
require credit on "every user interface screen", but rather speaks of a
splash screen (only) displayed "for sufficient duration to give
reasonable notice to the user of the identity of the Initial Developer"
-- with escapes to cover technology variations (e.g., no graphical
display) or if for various other reasons the requirement is
inapplicable.

Likewise as a very pleasant surprise, the MSPL's Exhibit B is worded in
a templated fashion, so that the licence can be reused for other works.

Last, the trademark clause, so excessive and problematic in other
Exhibit B examples, is perfectly reasonable:  Licensee acknowledges 
licensor's trademark title, and says that using them requires either
owner's permission _or_ any other entitlement created by law or by the
terms of the licence.

I find all of the above refreshingly reasonable.  The question you
asked is:  In my opinion, is it open source?

At the risk of being guilty of quibbling, a two-part answer:  (1)
Traditionally, restrictions on usage are seen as being inherently not
permissible in open source.  All restrictions.  This is seen as being
implied in a number of the Open Source Definition's provisions (e.g., #3
on derivative works, #6 on fields of endeavour, and #10 on technological
neutrality), even though it hasn't been called out as an explicit
prohibition in stark terms.  Strictly speaking, the requirement of a
splash screen in the graphical disply _if any_ is a restriction on use.

(2)  However, that restriction is of such a fleeting, minimal, and
reasonable nature that I think it should not raise objection.

The above is based on about an hour (only) of study and comparison --
and I have certainly made mistakes in such matters before.  However, I
am reasonably confident in my view.

Fair disclosure:  I was once employed at a firm then headed by Medsphere
former CEO and current Board member Larry Augustin, and knew him
personally for many years before that.  I esteem Mr. Augustin highly,
but hope and believe that my view of him personally has not coloured my
assessment of his firm's licence.

-- 
Cheers,                                      "Reality is not optional."
Rick Moen                                             -- Thomas Sowell
rick@linuxmafia.com